ABSTRACT End stage renal disease (ESRD) requiring hemodialysis (HD) affects >400,000 patients in the US, and over one million people will start dialysis in the next decade. While HD prevents immediate death due to uremia, patients continue to experience debilitating uremic symptoms including fatigue, pruritus, and anorexia among others. These symptoms are common?the prevalence of each is >30%, and >90% of patients have two or more?and are a major contributor to poor health-related quality of life (HRQOL). Thus, managing uremic symptoms is considered a top research priority by patients and physicians. However, the underlying cause of uremic symptoms is unknown. Our central hypothesis is that retained uremic metabolites, substances cleared by the kidney that accumulate in kidney failure, are the cause of uremic symptoms. Emerging metabolomics technologies provide an unprecedented opportunity to simultaneously assess thousands of blood metabolites in an efficient and unbiased fashion. Here, we seek to use a leading untargeted metabolomics platform to discover and validate the uremic toxins associated with uremic symptoms across two large hemodialysis cohorts that utilized a common, validated tool for symptom assessment. Our co-primary outcomes will be symptom scores for the three most common and debilitating uremic symptoms?fatigue, pruritus, and anorexia?each considered separately as their mechanisms may differ. Secondary outcomes will be: a) symptom scores for other uremic symptoms (nausea, difficulty concentrating, excessive daytime sleepiness, and pain) and b) HRQOL. For the discovery phase (Aim 1), we will identify novel metabolites associated with symptoms and HRQOL at study entry in the Longitudinal US/Canada Incident Dialysis Study (LUCID), an ongoing multicenter prospective cohort study of HD patients (N=700). We will consider metabolites individually, adjusting for multiple comparisons, as well as use systems-based approaches to account for metabolite inter-correlations and to assess results in the context of established biochemical pathways. Next, we will conduct internal validation (Aim 2) of metabolite associations with symptoms and HRQOL in a random sub-cohort (N=300) of LUCID at 1- year follow-up. For external validation (Aim 3), we will confirm these associations in a random sub-cohort (N=300) of the Hemodialysis (HEMO) Study, an NIH-funded multicenter randomized controlled trial of HD dose (Kt/Vurea) and dialysis membranes. This proposal brings together leading expertise in ESRD epidemiology, metabolomics, computational biology, uremic toxicity, and symptom science. If successful, our collective effort will provide valuable insight into the metabolic derangements responsible for uremic symptoms and poor HRQOL. Our findings will also spur mechanistic studies of uremic toxicity, advance the development of new treatments, and inform the design of patient-centered clinical trials, with downstream implications for patient management and health policy.